Plextor M5 Pro SSD Review @ [H]

FrgMstr

FrgMstr

Just Plain Mean
Staff member
Joined
May 18, 1997
Messages
55,634
Plextor M5 Pro SSD Review - Plextor brings the M5 Pro with the new Marvell Monet controller powering its banks of high performance Toshiba Toggle NAND. This is the debut of the first SSD with 19nm Toshiba Toggle NAND, the debutof the new Marvell Monet 88SS9187-BLD2 controller. We take a look and see what these new components bring to the table.
 
Will we ever get real world numbers in SSD testing ?

But still more ssd not poluted by anything in common with OCZ or Sand Force the better.
 
Michaelius said:
Will we ever get real world numbers in SSD testing ?

But still more ssd not poluted by anything in common with OCZ or Sand Force the better.
Click to expand...

What exact real world numbers would you like specifically?
 
LOL...

I love this sentence!

"With some of the sleep issues that have plagued a certain Marvell competitor this type of testing is crucial. "

Good job writing!
 
Am I supposed to take this review seriously without any mention/comparison to Samsung's new 840 SSDs? Christmas break came early or what guys? I mean you guys gave it a rating at the end which usually means something to me when I see it, but without any comparison to the top dog in the market, Samsung 840 Pro, I don't know what I'm supposed to take from the Silver rating. Sorry for being so critical.
 
The 840 hasnt been tested with any type of steady state that i have observed, and also they aren't releasing samples to even those reviewers that do only fill testing (that I am aware of).
By our measuring stick, the fresh results do not tell the real story. Actually by any measuring stick!
Of course we would love to do a review of the 840, and will see if we can procure one.
@AlienKing- good catch!
 
Plextor is the first to publicly use the new Marvell Monet 88SS9187-BLD2 controller. Thank You for the review and lol at the crucial bit :p
 
yes, OCZ rebrands the same controller with their own FW, which is a big reason I left the V4 in the comparison pool. Frankly I was surprised that they were still some similarities between the two. Seems that one of the major differences was OCZs implementation of the two modes, with the storage mode being slower.
 
Hugh_Briggs said:
The 840 hasnt been tested with any type of steady state that i have observed, and also they aren't releasing samples to even those reviewers that do only fill testing (that I am aware of).
By our measuring stick, the fresh results do not tell the real story. Actually by any measuring stick!
Of course we would love to do a review of the 840, and will see if we can procure one.
@AlienKing- good catch!
Click to expand...

Ahh, thanks for the clarification. Can't wait to see it when it comes out.
 
Hugh_Briggs said:
yes, OCZ rebrands the same controller with their own FW, which is a big reason I left the V4 in the comparison pool. Frankly I was surprised that they were still some similarities between the two. Seems that one of the major differences was OCZs implementation of the two modes, with the storage mode being slower.
Click to expand...

I thought the V4 used a 9145 controller? Has anyone published evidence of it using the 9187? I don't really have good information on this.

I am debating between the V4 and the M5P, and if the differences are NAND, firmware, and price... it may be tempting to go with the product which has the proven NAND, tested-by-more-owners firmware, and lower price.

But, Plextor's rep > OCZ's rep.
 
MrAgmoore said:
Samsung 840 Pro vs. Plextor M5 Pro

http://www.anandtech.com/bench/Product/665?vs=646


I have no idea which stats are the most important for a gaming SSD.
Click to expand...

Sequntial read/write and 4K. Anyway, the only true performance drive with sustained read/write performance (on heavy load) atm is 840 pro, it scores an insane 40mb/s higher results then M5 pro in Anands heavy workload which is typical usage scenario when you reformat your drive, and install tons of stuff in one day.
 
From what I understand, random 4k is the most important metric for OS and probably for many games too, having to access a bunch of small files. A typical consumer hard drive will struggle to get even 1 MB/s read speeds, whereas a modern SSD is 100+ times faster.

Also, while sequential speed helps in theory, I think the bottleneck to many games may be elsewhere. For instance, Shogun 2 has legendary load times when maxed out on DX11 and all the optional settings, and it apparently is CPU-bound due to the DX11 textures. I switched from a M4 to 830 (both 256GB) and found no appreciable difference in loading time. I switched to DX9 and loading times became bearable.

I'm on a 3570K@stock + 7970@stock which isn't exactly slow.. nor is my RAM slow (16GB of 1866MHz, CAS9).

P.S. I refuse to buy any more Total War games until they make use of more than one thread. It's 2012 and they are still using #*%)ing single-threaded b.s. No more, Creative Assembly, no more.
 
Last edited:
There are many bottlenecks to gaming, and storage is one. There is no single improvement that will boost your game loading times more than an SSD.
BUT the advantages in-game are whats important, and there are huge advantages in gaming to be had with an SSD.
 
Here are the benchmarks for my PX-256M5P on P8Z77-V PRO.

With write-caching:
cdm1yeswritecaching.png

Uploaded with ImageShack.us

Without write-caching:
cdm1nowritecaching.png

Uploaded with ImageShack.us
 
So, the new 1.02 firmware update for the M5 Pro is available, advertising random-reads of up to 100,000 IOPS.

The firmware can be downloaded on the M5 Pro firmware page.

I tested this upgrade on my 128gb M5 Pro, and, well... pretty much nothing:

Before:
WHlTf.png


After:
O2tkq.png


As you can see, not really any difference. In fact, the first test I ran clocked in at 940 total, a couple of points less than the 1.01 firmware results.

tl;dr: Feel free to grab the update, but don't expect wonders (at least on the PX-128M5P). YMMV.
 
LifesBane said:
So, the new 1.02 firmware update for the M5 Pro is available, advertising random-reads of up to 100,000 IOPS.

The firmware can be downloaded on the M5 Pro firmware page.

I tested this upgrade on my 128gb M5 Pro, and, well... pretty much nothing:

As you can see, not really any difference. In fact, the first test I ran clocked in at 940 total, a couple of points less than the 1.01 firmware results.

tl;dr: Feel free to grab the update, but don't expect wonders (at least on the PX-128M5P). YMMV.
Click to expand...

Just updated and the maximum IOPS definitely increased for reads like they said they would (93K with 1.01 vs 98K with 1.02). That's all they really said was coming with this firmware release.

I'll probably edit this post with a new score after my system idles overnight and keep in mind these are with a laptop :p

m5p.jpg


Edit: After logged out/idling overnight

m5px.jpg


JoeComp said:
The main difference with the new M5P firmware is that 4KQD32 reads go up from 94K IOPS to 100K IOPS, but only on the 256GB and 512GB models. The 128GB model remains about the same. Even on the 256GB and 512GB models, almost nobody will notice a difference, since 4KQD32 reads only matter for workloads that are extremely unusual among consumers.
Click to expand...

This^
 
Last edited:
The main difference with the new M5P firmware is that 4KQD32 reads go up from 94K IOPS to 100K IOPS, but only on the 256GB and 512GB models. The 128GB model remains about the same. Even on the 256GB and 512GB models, almost nobody will notice a difference, since 4KQD32 reads only matter for workloads that are extremely unusual among consumers.
 
You must log in or register to reply here.
Back
Top